Traditional chiral separation membranes, such as those composed of linear chiral polymers, often struggle to simultaneously achieve high enantioselectivity and high permeability due to issues like low density of recognition sites and poor structural stability, facing the significant “selectivity-permeability” trade-off. In this study, a three-dimensional cross-linked chiral composite membrane (P-SiO2) was constructed on a porous silica (SiO2) support via surface-initiated polymerization, employing a thiol–ene click chemistry strategy. This 3D cross-linked network not only enhanced the spatial regularity and interfacial stability of the chiral recognition sites but also endowed the membrane with a network flexibility intermediate between that of linear polymers and rigid porous materials, effectively improving the membrane’s separation performance and mechanical strength. The effects of key parameters, including the type of cross-linker, the pore size of the SiO2 support, reaction concentration, and the content of the chiral monomer, on the membrane structure and separation performance were systematically investigated. The results demonstrated that the P-SiO2 composite membrane with cross-linked network exhibited excellent chiral resolution performance for the d,l-phenylalanine racemate, with a permeability coefficient of 1.97 × 10–10 m2/h and an enantiomeric excess (ee%) as high as 98.3%. Mechanistic studies revealed that the chiral recognition process follows a delayed transport mechanism, in which the composite membrane exhibited stronger adsorption for l-phenylalanine, thereby delaying its transport and allowing the d-enantiomer to permeate preferentially. This study provides an innovative design strategy and a feasible preparation pathway for developing chiral separation membrane materials that combine excellent film-forming ability, high enantioselectivity, superior stability, and good mechanical properties, showing broad application prospects in the efficient separation and purification of chiral pharmaceuticals.
Zang et al. (Thu,) studied this question.